opening quote after a modal verb

What this neuron does

This neuron lights up on an opening quotation mark that comes right after words like "can", "could" or "may", where a word or phrase is being put in quotes for effect. When it was turned up, the model's writing drifted into rambling and unrelated topics, so the quote pattern is the only clear link.

Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens a prefilled issue on GitHub).

Fires most on

The text windows where this neuron's activation was highest. Amber marks positive activation; the outlined token is the peak. Hover a token for its strength.

1. #1 peak

   Their study, which is published in the journal Proceedings of the National Academy of Sciences (PNAS), is the first to identify a non-primate species as having “orthographic” abilities.⏎In the experiment, pigeons were trained to peck

2. #2 peak

   .⏎The combined effect was that Myotragus was sluggish, with slow reaction times, the bone study suggests.⏎Like modern-day reptiles, the goats probably “saved as much energy as possible just lying around and basking in the sun,”

3. #3 peak

   to 100⏎tspan=[t0 tend]; %need to use it in this form in MATLAB 5.⏎[t,y]=ode45('rhs',tspan,y0)%call the routine ode45⏎plot(t

4. #4 peak

   BUQUERQUE, NM – A team of researchers from Sandia, Lawrence Berkeley and Pacific Northwest National Laboratories tested a “sponge-like” mineral that can “absorb” uranium at a former uranium

5. #5 peak

   Health Checklist⏎Being a college student warrants a quick health check in order for you to remain healthy all school year. By following a few simple strategies, you can “rock on” in school, work and your social life.⏎Get involved in regular

6. #6 peak

   landing, as well as what it is like to do a spacewalk<|endoftext|>Global warming “can be reversed", scientists claim⏎Hi-tech new bio-energy plants could “reverse” global warming by pumping carbon dioxide into old gas wells - lowering temperatures by

Steering

Push this neuron up or down while the model writes, and watch what changes. The outputs were generated ahead of time, so nothing runs in your browser.

Most negative on

Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.

1. #1 peak

   ', 'pomegranite', 'bananana', 'grape', 'loganberry'] print "You are an" , random.choice (verb) , random.choice (fruit)⏎Any Tutorial Videos?<|endoftext|>What is globalisation?⏎Global

2. #2 peak

   Social Cognition in Non-human Primates Using a Network of Computerized Automated Learning Device (ALDM) Test Systems⏎Institutions: Aix-Marseille University.⏎Fag

3. #3 peak

   that will give a useful input to those interested in Artificial Intelligence in education.⏎|Subjects:||L Education > L Education (General)|⏎|Deposited By:||Zaidatun Tasir|⏎|Deposited On:||

4. #4 peak

   structure = function.⏎(Image from: https://commons.wikimedia.org/w/index.php?cur

Which neurons in earlier layers feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.

Activation histogram

How often this neuron fires at each strength. Almost always it does close to nothing (the tall bars in the middle); the snippets above are the rare moments it fired hard, amber for positive and blue for negative.

Logit lens

The words this neuron nudges the model toward (amber) or away from (blue) when it fires. A hint, not a label: many neurons work through later layers instead.